Shifting Perspectives: The sun and the moon, page 3
by Tyler Caraway 
Aug 27th 2010 at 5:00PM
I speak often of high-end raiding content, though to ignore the a significant portion of the game is rather foolhardy. Leveling and solo content is a significant portion of this game, and it is one that should not be overlooked. I have briefly touched upon this subject previously, how Eclipse is not the best of concepts for leveling, but I would like to take the time to expand upon those initial thoughts.
I have spent the past few days getting a feel for balance leveling, predominately without the use of heirloom items. I also spent some time testing it against feral leveling to see how the two compare -- the basics of which are that there is no comparison. That is a different story, for now I wish to focus exclusively upon the concept of Eclipse. As far as leveling is concerned, Eclipse may as well not exist. Without haste -- and to a certain extent, crit -- you simply cannot reliably generate Eclipse fast enough to actually get a proc. Due to the rapid out-of-combat decay that Eclipse experiences, it often requires three mob kills just to get a proc, though that all depends on how spaced out the mobs are and how quickly you can re-enter combat. Worse than that, even once a proc is gained, Eclipse still rapidly decays out of combat, and a vast majority of the proc will generally be lost before it can even be used.
Even after the changes brought forth in Cataclysm to Eclipse and the way in which our damage is balanced, Eclipse is still quite a significant portion of our overall potential. The inability to fully or even moderately benefit from Eclipse while leveling is crippling to a young balance druid. A 25 percent damage boost is significant, even at lower levels. To lose that slows the entire process drastically. Sadly, there is no real means to fix this situation; it is merely a byproduct of the Eclipse system that cannot be helped. Reducing the rate a which Eclipse decays while outside of combat would certainly help the matter, but it wouldn't entirely solve the issue.
Eclipse simply is not a fantastic leveling concept, and I believe that is merely something that balance druids are going to have to live with. The mechanics are solely developed around factors of prolonged combat; they simply aren't set up to exist within the confines of short periods of combat. We can work with that, but solo leveling may need a spot of help. That doesn't have to come from Eclipse; the ability doesn't need to be tweaked in any significant means in order to assist the leveling process. If balance leveling does need help (and I believe that it does), then there are other means to do that.
Closing thoughts
I realize that I have rambled on for quite a while, and I do not wish to do so for much longer. I simply would like to restate that I do not hate the current way Eclipse works on beta. It is mathematical and predictable, but that in of itself is not a negative quality. While I make arguments for a more complex rotation, I do so out a pure desire to have such a thing, not out of a need for it. Whether or not the balance rotation becomes more complex is of no consequence. It would be nice to have, I would say, but it isn't a requirement. Far more important than that is the sustainability of our DPS potential; that we are able to scale appropriately as the expansion progresses. To that end, there are some potential changes to Eclipse that should be made in order to ensure that. While I cannot say with 100 percent certainty that it has to happen, changing Euphoria to act as a "smart" Eclipse gain would make the balancing act of our DPS scaling far easier to control.
With Euphoria as it remains, how we scale is then subject to fairly vast fluxuations, which can cause long-term issues much as we experience currently. Scaling does not need to be linear, but it should be far more controlled than the way it currently sets out to be. Scaling is important; steady, more predictable scaling, far more so. If a talent, ability or mechanic scales too well or too little, it becomes far easier to adjust the parameters that need to be changed when the scaling itself is more controlled. When the scaling is far more chaotic, the proper adjustments become much more difficult. Euphoria must be changed, if not for the sake of allowing our DPS to be balanced properly, then for the purpose of allowing our DPS to be easier to adjust, should it be necessary.
Every week, Shifting Perspectives treks across Azeroth in pursuit of truth, beauty and insight concerning the druid class. Sometimes it finds the latter, or something good enough for government work. Whether you're a bear, cat, moonkin, tree or stuck in caster form, we've got the skinny on druid changes in patch 3.3, a look at the disappearance of the bear tank and thoughts on why you should be playing the class (or why not).
Reader Comments (Page 1 of 3)
Deathknighty Aug 27th 2010 5:16PM
I'm pretty sure I didn't miss anything (forgive me if I did), but I still have no idea what you mean by a "smart" Eclipse gain.
How would that work, what would the mechanic be? :)
Tyler Caraway Aug 27th 2010 5:34PM
It's possible that I didn't explain it well enough. Here's what I mean:
When you enter into combat, the Eclipse bar will usually be at the middle or 0 point. From there you need to either reach -100 or +100 in order to proc Eclipse. Every time you cast Wrath, you move the scale by -13. Every time you cast Starfire, you move the scale by +20. This never changes -- Wrath will always move it -13 and Starfire always +20 without exception.
Starsurge is different. When you cast Starsurge, it can move the scale by either -16 or +16 depending on where you are on the scale. Perhaps the best way to describe this is to say that Starsurge always moves the scale in the same direction that it is already at. If you cast a single Wrath and are now at -13, then Starsurge will move the scale by -16. If you move the scale all the way to -100, thus getting a Lunar Eclipse, and then cast a single Starfire (which is +20) you would now be at -80 on the scale. If you then cast Starsurge, it would move the bar by -16 still since you are at the negative side of the scale. If you cast Starsurge while at +80, then it would move the scale by +16 since you are on the positive side of the scale.
Essentially, I want Euphoria to work the same way. Instead of always being either a - or a + depending on what spell you cast; I'd much rather it change based upon where you stand on the Eclipse bar. If you are on the - side of the bar, then Euphoria procs should provide a - gain of Eclipse, if you are on the + side of the bar, then Euphoria procs should provide a + gain of Eclipse.
Hopefully that explains it a little better. :\
Deathknighty Aug 27th 2010 5:36PM
Ok, thanks. :)
talitha3k Aug 27th 2010 5:16PM
sigh. I got my first taste of the new Eclipse mechanics last night playing the beta. I am not at all impressed. in fact, I HATE it. I do not see what was wrong with the RNGness of it now.
Lissanna Aug 27th 2010 7:38PM
because it makes us really great DPS when we stand sill and are lucky, and really bad DPS when we have to move and are unlucky. It makes random have a bigger impact on our DPS than skill.
Zaros Aug 28th 2010 12:47AM
All caster classes lose DPS when they move. The problem was that if we had to lose we lost DPS. making it last 45 seconds but be consumed in less then that time was a fix for the movement problems but it removed our only proc that really changed our rotation.
Every proc we have (besides eclipse) don't change our play style and while useful just don't add to the game-play.
Eclipse is what made our rotation interesting and now we need something to fill the empty void in our hearts.
Mind you, I love the new system. but we have nothing left anymore.
Just wondering does natures grace still scale badly with wrath? My natures grace hasn't been working but a GM is "working on it". Anyone else agree that they need to redesign natures grace?
Tyler Caraway Aug 28th 2010 7:54AM
Wrath has a base line 2.5 second cast now, so 2 second talent, and Nature's Grace is down to 15% haste instead of 20%. The only feasible way in which you'll be able to GCD cap Wrath would be if you somehow managed to get 50% haste, and then you would only hit the GCD on Heroism.
Sanat Aug 27th 2010 6:48PM
From a pure mathematical stand point crit shouldn't affect the long term gains from the mastery stat. This is because currently with increased crit you gain the buff faster, but you also burn it faster. While I agree having a smart system would be great across a massive data pool you should see crit and mastery act independent of each other as a DPS increase.
This is assuming I have read everything correct. Tyler is this correct?
Snilloc Aug 29th 2010 7:20AM
From what I've read, it seems that the negative scaling with crit is due to the fact that with a specific number of crits the ratio between spells cast during the eclipse and those cast to proc the next eclipse will decrease meaning that we benefit less from mastery. This is because the last spell you cast before the eclipse proc ends benefits from the full value of eclipse regardless of how far past 0 it goes. By making euphoria generate eclipse power based on our position on the eclipse bar, crit would never decrease eclipse uptime.
ambermist Aug 27th 2010 6:06PM
I wasn't thrilled with the new Eclipse as I quested through Hyjal. You're right about it not being a good leveling tool, and that exists whether you're leveling at 30 or 80. However, once I hit an instance (in this case, Blackrock Caverns) and got a feel for it on a longer fight, I don't hate it.
The one thing that's stumping me is exactly what to do with Starsurge right now. I am not good with numbers, period. I have to trust that the theorycrafter consensus is going to be accurate enough to remain competitive dps as long as long as I execute correctly.
I haven't found a consensus on Starsurge (although admittedly I haven't spent a lot of time looking for one). Right now it feels like the third wheel in my rotation. Is it better to use it to get to an Eclipse? Should I use it to suspend Eclipse a cast or two longer? Or should I use it on cooldown? So far I've been doing a combination of the first two, using it to get to Eclipse if the boss still has a lot of health, or using it to maintain an Eclipse when I have a good proc or when the boss is almost dead and one or two casts will finish it.
I'm also a little lost on the spellpower to intellect conversion and how that should or shouldn't change my gemming and enchanting, but that's a topic for a different post.
Tyler Caraway Aug 27th 2010 7:43PM
Right now, the best thing to use Starsurge for is during an Eclipse proc. This is done for two reasons:
A) Starsurge gains a damage benefit from both procs
B) Starsurge increases the duration of Eclipse (effective duration anyway.)
It's likely that Starsurge may come off cooldown while going inbetween procs, but if it won't be up in time to use during the next proc, then it is best to just save it until that time. Even though Starsurge is our highest damaging spell, the fact that it's currently the only means of getting a double benefit from Eclipse is what really causes it to shine.
ambermist Aug 27th 2010 8:13PM
Thanks, that's exactly what I wanted to know. :-)
Matthew Aug 27th 2010 6:54PM
Here's what I'd like to see for leveling or PvP:
a Warhammer-inspired eclipse:
Heals generate DPS potential, DPS generate Heal potential.
Throw some hots, then nuke for a while, then back to hotting.
Alas, Restokin is not considered a 'spec'. Thus spoke Zoroaster.
Thanks for the thoughtful article, Mr C.
Lissanna Aug 27th 2010 7:39PM
At this point, Cata has killed off any hopes I had of restokin talent builds coming back. :(
Tooay Aug 27th 2010 6:56PM
Some of the issues you mention, namely mastery only ending up affecting some spells or having eclipse up for only 55 to 65% of casts, are only real problems if druids aren't balanced around these numbers. It's not an issue to have only let's say half, maybe even less of your spells benefit from a key buff as long as the numbers are arranged in a way that half uptime is the intended uptime.
Take frost mages; you need to cast Frostbolt to gain the mastery buff, which is consumed on another spellcast. That means at most, you'll benefit from it 50% of the time, but then if you do consume every other cast, you'll probably be casting spells that amount to a dps loss. In the end though that's not an issue because the deathfrost buff will, once balanced, be significant enough that this occasional bonus makes the difference it needs to.
It also reminds me of the shaman concern about their mastery (more healing done at lower health). Some were worried that their numbers would be uncompetitive unless they kept their targets under 50% health, which could be dangerous, and will get sniped by other healers anyway. Ghostcrawler eventually came to the forums and said that shaman's healing will not be balanced around shaman healing target below 30% health; it'd just be an extra effectiveness boost when they do. The same will probably be said for druids; a slightly lower uptime for eclipse will not be dramatic because druids most likely will be balanced around the -minimum- uptime, and any extra duration is just icing.
Tyler Caraway Aug 27th 2010 8:09PM
Yes, I am no all too concerned about uptime in the sense of balance in terms of our damage potential, rather in terms of scaling potential. While our damage is a primary concern, ensure that this new stat that Blizzard is spending time creating is at least halfway viable is equally important. We don't need another repeat of the current spirit mechanic wherein mastery, for balance druid at least, as a gear stat ends up being half the value of that of crit or haste, thus causing us to avoid it like the plague.
Balancing around proc uptime, however, is a very tricky subject; more so when you get into variable uptimes. In our current state, our balancing -- for everything that's been done with it -- is pretty bad. Erdluf postulated that we currently see around a 76% benefit from Eclipse, and I'd contest that it's actually a bit higher than that for the higher end balance druids. In fact, it's really something of a disastrous issue right now. Just speaking about personal data -- with an average Eclipse spread, I place around average within the DPS ranking of my guildmates. If you happen to be going for the Been Waiting a Long Time for This achievement, then phase 1 of Lich King is an excellent tool for taking a snapshot in balancing (not 100% accurate, mind, but it's the only Patchwerk-esque encounter you'll find (Rogues and Hunter data will certainly be skewed to the lower end here, but Warrior, Shaman, and other cleave based data won't be))
Anyway, as I said, with an average Eclipse proc rate, my DPS potential is probably within the range of where it should be all things considered -- I'm middle of the pack, not great, not bad, just there. Given how heavy we run on pure DPS, that's a good spot to be in. However, with exceptionally good Eclipse proc rates, my DPS can soar far and above the given potential of any class. There have been numerous times where I've merely been utterly lucky with Eclipse and blown every other DPS out of the water on any given encounters; it's all a matter of dice.
That's mostly due to the current harmful aspect of RNG that Eclipse suffers from -- the new Eclipse holds the reverse in that RNG will have virtually no impact on DPS (at least as far as Eclipse is concerned.) Euphoria as a talent is an utter wash -- having it or not having it will not have any noticeable impact on DPS that wouldn't be attributed that any standard RNG variation. It currently sits as a RNG element that holds no positive aspect and is nothing more than a potential ticking time bomb.
As mastery increases, Eclipse uptime becomes more important, and as crit increases, the variable rate of uptime has a wider breath of changing. We simply do not know at this time how these two aspects are going to scale in relation to each other, all we can say is that the data shows Euphoria holds the potential to either completely screw the druid, or to be a very modest boost in DPS -- think Genesis levels. RNG can only be balanced around to a generalized sense, one can never pinpoint balance every aspect of any given spec so long as RNG exists. What we can do is control the deviation.
Right now, Euphoria stands to hold either a positive or a negative deviation -- however minute it might be. A talent should never, ever have the potential to lower a player's DPS no matter how small a chance that might be. This is what we currently face with Euphoria. Changing Euphoria to a "smart" system avoids that completely -- there is never a change for a negative deviation, only a positive one. Is the talent itself currently balanced for such a change? Potentially not, more testing and access to actual gear and ratings is needed before that assumption can be made, but it is an easy fix either way. If changing Euphoria to a "smart" system makes the talent too powerful, then simply reduce the energy gains. Or you can reduce how potent the mastery stat is.
In any circumstance, this type of change won't have that drastic of an effect on Eclipse overall. All it would mean is that, for every other critical strike you have, you get an additional spell cast under Eclipse, and there's a set maximum of casts that you'll be able to reach. The RNG aspect still exists, but the are set limitation to control it, isolate any potential issues, and correct them. The current system simply does not afford the same flexibility that a "smart" system would in terms of balancing.
mistairdvant Aug 28th 2010 4:26AM
So I can't Wrath spam anymore, right?
Pyromelter Aug 27th 2010 7:00PM
Murmur's, I love how thorough you are with Moonkin.
It's too much though man. I don't think you've written one Shifting Perspective that has been just one page. I know the Moonkin community generally gets really deep into the numbers, and that's great. I have a hard time getting through your blogs man... it reads more like a technical manual than a weekly editorial.
I feel like there is a lot of effluvia you put into your articles, and that it would be 1000% more readable if you cut out a lot of redundancies, or at least tightened your focus a bit.
I'm not going to sit here and be an armchair editor or grammar Nazi. But I will give examples of what you should strive to stop doing.
-Phrases like "Clearly," "Obviously," "All of that aside" (page 2); "Clearly," "You see" (page 1)
-A statement like "Take note that I am not stating this as a negative, merely as the fact that it is." (page 2) - A statement you make will be taken as negative, positive, or neutral, no matter what you tell us you think it is. It is completely a waste of space to make such a comment
And what you should strive to do:
-A one page article. Just one. See if you can do it. I know you can brother!
-Maybe a little more fun outside of the numbers? No one is going to expect you to be Matt Rossi or Christian Belt, but we're big furry lazerchicken and make big boom. A little more fun, like instead of saying "allowing our DPS to be easier to adjust," make it like "allowing our DPS to have a lot more boom boom pew pew lazors if we're falling behind" Well okay can't see you saying that, but hopefully you get my drift.
-Tighten the focus. Yes, eclipse is big for moonkin, and it's very mathy and important and all these thing affect it. You can still tighten it up and keep the focus so that you aren't running into 3 pages on blogsmith's blogger pages.
And apologies for being the armchair editor (even though I said I wasn't). It's just gotten to the point where I felt like I had to say something.
cui_h_1999 Aug 27th 2010 8:12PM
This.
And I'm a raiding fatkin w/ 10/12 hm exp (Sina/LK on call). I would love to hear more about gimmicks to squeeze DPS from fights, fatkin soloing, and discussion about trinkets rather than a block of text about Eclipse (since Blizz won't be changing it till Cata anyways).
Tyler Caraway Aug 27th 2010 8:27PM
Out of my last 6 SP articles, 5 of those were 1 page long, and 1 of them was 2 pages. Prior to that, all of them were a least 2 pages long, with very few 1 page articles. I've been trying to cut down on volume quite a bit, but it just doesn't sit all that well with me. I know that the longer articles aren't always better and brevity is the essence of wit, but there is simply so much information that I can put out there and I am compelled to do as much as I can. For as long as these articles are, few people realize how much time I really do spend on trimming them down -- as an example, I cut out a solid 2,000 words from this article before submitting it. I do try to be concise, but sometimes I do deal with somewhat complex subjects and I'd rather be accused of being long winded than being unclear on any topic.
I've tried fun in the past, it generally hasn't gone over all that well. ;) I'm simply not Rossi, nor Allison, nor one of the slew of other very talented writers for this site. I operate with bland subjects, I try to make them as appealing as I can, but, at the end of the day, I'm just a simple theorycrafter that's trying to entertain people with numbers as my only tool.I am trying, believe me, but my humor is limited solely to sarcasm, which simply doesn't translate all too well in this setting -- or most virtual settings for that matter.
As for focus, it is as tight as I can get it given the circumstances. I do have a tendency to ramble off topic on subjects I get excited about -- and balance druids is one of my favorite subjects -- but I do work to limit that here. The leveling issues of Eclipse, for example, could easily have been expanded into it's own two, or, yes, even three page article. You see, though, this places me in a bit of a conundrum. I could have simply cut the leveling aspect out of the article entirely and made it into a separate one, but then that's an additional week were I'm writing about the same dried out, worn down topic that people are likely to get bored with. I've already done far more articles on Eclipse that I'd like to for Cataclysm and I don't want Shifting Perspectives to turn into "Hey, look, the weekly Eclipse article is up again!" To avoid that, it does mean that other articles are going to have to run longer so that I can cram more and more information into them. That's another thing I'd rather be accused of -- to me, it is better to be considered wordy that to be thought of as always running the same content over and over and over again.
You may also arm chair edit me any time you wish. Fair warning though, another of my flaws is that I do enjoy arguing for the sake of it most of the time. ;)